Graphics virtualization may enable software executing within a virtual machine (VM) to control various events and have access to graphics hardware resources on a physical machine, wherein a virtual machine monitor (VMM) may create and run the VMs on the physical machine. Accordingly, the VM software may collectively be referred to as guest software and the VMM may be referred to as a host. Partitioning the graphics hardware resources between multiple VMs may present certain challenges with regard to efficiency and security. For example, the host and the guest software in conventional graphics virtualization solutions may not have the same view (e.g., size and layout) of the graphics memory address space, nor can the guest graphics memory address space be identical to the system memory address space in certain instances. Accordingly, as guest commands (e.g., rendering commands) are issued by the guest software to the graphics hardware, address remapping and/or fixing (e.g., finding the guest address in the command and replacing it with a host address) from the guest view to the host view may be conducted. Hardware based address remapping and/or fixing may involve tagging technology that adds complexity. Moreover, software based address remapping and/or fixing may involve VMM interception, parsing and/or conversion of each guest command, which may cause additional overhead.